The RSI ... modified

[Trader.Fighter]

Quote from HurricaneUS:

Why would you trade anything else besides a trend

You can always find some market that's in a good uptrend/downtrend...even better if you can combine the indicators with fundamentals based price targets and read of institutional buying activity...but I digress

Could not agree with your more, just felt it was a necessary addition to your previous commentary.

Thank you

 

[carlberky]

Quote from MAESTRO:

Oh, brother .... THERE ARE NO SUCH THINGS AS PREDICTIVE INDICTORS THAT ARE BASED ON HISTORICAL DATA ALONE!!!!!

Unicorns in the enchanted forest, clouds in the sky that look like dog heads and the desert mirage that looks like an oasis. That is all to it.

Cheers,

MAESTRO

If you do not see "some predictive value" in the attached segment, then you looking but not seeing.

 

[maxpi]

Quote from MAESTRO:

Oh, brother .... THERE ARE NO SUCH THINGS AS PREDICTIVE INDICTORS THAT ARE BASED ON HISTORICAL DATA ALONE!!!!!

Unicorns in the enchanted forest, clouds in the sky that look like dog heads and the desert mirage that looks like an oasis. That is all to it.

Cheers,

MAESTRO

It's impossible and Maestro will prove it!

 

[Sergio77]

If there is a trend then buy a hold and possibly pyramid if you are brave. If there is no trend stay out. This is how big money is made. All else, including pseudo-predictive indicators, mean-reversion, pairs trading, statistical arbitrage and so on is for suckers.

 

[trilogic]

Quote from Sergio77:

If there is a trend then buy a hold and possibly pyramid if you are brave. If there is no trend stay out. This is how big money is made. All else, including pseudo-predictive indicators, mean-reversion, pairs trading, statistical arbitrage and so on is for suckers.

looking at a daily chart- what is a "trend" how to measure ?

 

[jsp326]

Quote from carlberky:

Another interesting discovery was that there seems to be some predictive value, with the indicator peaks and valleys coming before or with the price. A quick glance at other commodity charts had similiar results.

If predictive value was determined by "quick glances" at charts, all traders would be rich.

 

[carlberky]

Quote from jsp326:

If predictive value was determined by "quick glances" at charts, all traders would be rich.

The "quick glance" was not for trading, but to varify that "the indicator peaks and valleys coming before or with the price" was not unique to the chart shown.

 

[jsp326]

Quote from carlberky:

The "quick glance" was not for trading, but to varify that "the indicator peaks and valleys coming before or with the price" was not unique to the chart shown.

That still doesn't prove anything. Design a set of objective trading rules, backtest test it over a number of years with hundreds of trades, then see if you have anything worthwhile.

 

[carlberky]

Quote from jsp326:

That still doesn't prove anything. Design a set of objective trading rules, backtest test it over a number of years with hundreds of trades, then see if you have anything worthwhile.

If I did that, would it "prove anything" to the predictive value skeptics on this forum?

 

[jsp326]

Quote from carlberky:

If I did that, would it "prove anything" to the predictive value skeptics on this forum?

Assuming you did it properly -- and it was robust and profitable -- then yes. There are still a lot of potential pitfalls with backtesting, but it would be much better than screenshots from charts. I've seen tons of indicators/systems that look good on a variety of charts but fail to be profitable on long-term tests.